- 18 11月, 2020 2 次提交
- 17 11月, 2020 6 次提交
-
-
由 Huihuang Zheng 提交于
Fix a bug that used in PaddleGAN model which used `isinstance` in a for loop
-
由 Jacek Czaja 提交于
-
由 wangchaochaohu 提交于
-
由 littletomatodonkey 提交于
* adapt pad const * fix comment and rm fluid import * rm stdout * fix note
-
由 Aurelius84 提交于
* fix output dtyp inconsistent with input * refine code
-
由 zhupengyang 提交于
-
- 16 11月, 2020 11 次提交
-
-
由 Chen Weihang 提交于
-
由 cc 提交于
-
由 wawltor 提交于
fix the gradient bug for the topk v2
-
由 Kaipeng Deng 提交于
* add WeightedRandomSampler. test=develop
-
由 joanna.wozna.intel 提交于
* Add checkpoint to quantize * Change bfloat16 option
-
由 lidanqing 提交于
* add mkldnn inplace op version test * update mkldnn_inplace fuse pass * update the inplace test
-
由 pangyoki 提交于
* fix test_multinomial * fix test_multinomial add 0 prob
-
由 Kaipeng Deng 提交于
* DataLoader support not auto collate batch. test=develop
-
由 Huihuang Zheng 提交于
PaddleSeg uses tuple as parameter of transpose in dygraph code: https://github.com/PaddlePaddle/PaddleSeg/blob/release/v0.7.0/dygraph/paddleseg/models/danet.py#L152 However, in dy2stat, static code doesn't support the perm as a tuple. This PR fixed it.
-
由 Guo Sheng 提交于
* Fix gradients with ignore_idx in softmax_with_cross_entropy. test=develop * Fix gradients with ignore_idx in softmax_with_cross_entropy on cpu. Remove softmax_with_cross_entropy from op_threshold_white_list. test=develop * Fix test_softmax_cross_entropy_op.py. test=develop
-
由 YUNSHEN XIE 提交于
-
- 13 11月, 2020 4 次提交
-
-
由 Huihuang Zheng 提交于
I found the unittest failed due to batch size. Maybe the reason is that our CI machine has limited memory. I decreased the batch size.
-
由 lilong12 提交于
* update, test=develop
-
由 Yang Zhang 提交于
this is a preemptive measure, the unit test was failing occasionally with cudnn internal error
-
由 channings 提交于
* add onnx export module, test=develop * add unit test for paddle.onnx.export * adjust api & doc * fix some typo
-
- 12 11月, 2020 4 次提交
-
-
由 Yang Zhang 提交于
also fix `no_grad` alias
-
由 Shang Zhizhou 提交于
* skip_layernorm_op done * add unittest * slice op convertor support trt < 6 * skip_layernorm only work in ernie
-
由 joejiong 提交于
As the title
-
由 lidanqing 提交于
-
- 11 11月, 2020 4 次提交
-
-
由 Leo Chen 提交于
-
由 YUNSHEN XIE 提交于
-
由 liym27 提交于
1. Remove 'fluid'; 2. Variable -> Tensor APIs: sum, convert_call, convert_ifelse, convert_logical_and, convert_logical_or, convert_logical_not
-
由 Huihuang Zheng 提交于
Modify ProgramTranslator and TracedLayer Doc for API 2.0
-
- 10 11月, 2020 1 次提交
-
-
由 Huihuang Zheng 提交于
Remove "fluid", "variable" in 2.0 API doc
-
- 09 11月, 2020 8 次提交
-
-
由 YUNSHEN XIE 提交于
-
由 Huihuang Zheng 提交于
As the title
-
由 YUNSHEN XIE 提交于
* fix ut exec timeout notest,test=kunlun * fix error for executing ut timeout,test=document_fix
-
由 YUNSHEN XIE 提交于
* modified timeout value on windows * fix some error
-
由 YUNSHEN XIE 提交于
-
由 YUNSHEN XIE 提交于
-
由 Tao Luo 提交于
-
由 YUNSHEN XIE 提交于
* exec ut no more than 15s 2 * fix for ut test_inplace_addto_strategy timeout
-